Dehydroepiandrosterone (DHEA) supplementation results in supraphysiologic DHEA-S serum levels and progesterone assay interference that may impact clinical management in IVF

Abstract

Purpose

Dehydroepiandrosterone (DHEA) is often prescribed for poor responders in IVF in an effort to improve response to ovarian stimulation. The effect of DHEA supplementation and resultant supraphysiologic DHEA-S serum levels on sex steroid assays has not been evaluated in this population. This study seeks to determine the relationship between DHEA supplementation and progesterone measurements to characterize the degree of interference with particular immunoassays.

Methods

Characterization was accomplished in two phases. First, DHEA-S standard control reagents with no progesterone present were assayed for both DHEA-S and progesterone levels. Second, serum pools from 60 unique IVF patients’ serum were used to create six pooled serum samples: three from patients on DHEA supplementation and three from patients not on DHEA supplementation. The three pools were composed of patients whose serum fell into low, medium, and high progesterone ranges. Baseline DHEA-S and progesterone were measured, and the mean level of DHEA-S in the mid-range progesterone pool was used as the mid-point for addition of DHEA-S standard to the serum pools from patients without DHEA supplementation. Progesterone from these pools was then measured on three commercially available immunoassay systems.

Results

The first experiment revealed a linear increase in progesterone when analyzing the DHEA-S standard ranging from 0.5 ng/mL in the blank control (no DHEA-S) to up to 2.0 ng/mL in the high control (DHEA-S >700 μg/mL), indicating that the DHEA-S cross-reacts with the progesterone assays. In the second experiment, patients’ serum DHEA-S and progesterone were measured from pooled serum samples of those taking DHEA and those not taking DHEA. Adding DHEA-S to the pooled serum of those not taking DHEA resulted in a linear increase in progesterone levels on two of three commercially available immunoassays (p < 0.05).

Conclusions

DHEA-S can interfere with standard progesterone immunoassays used in clinical ART programs, and thus serum progesterone levels in IVF patients on DHEA supplementation may not reflect truly bioactive progesterone.

Introduction

Dehydroepiandrosterone (DHEA), an adrenal androgen, is being increasingly used for poor responders in an attempt to improve response to stimulation during IVF treatment [1]. The first case series report was in 2000 and demonstrated increased response as measured by the estradiol/ampule of gonadotropin utilized; however, very little class I data exists to support or refute its use in this setting [2]. Nonetheless, survey data exists to show that over one quarter of IVF practitioners adds DHEA supplementation to the regimen in those with poor ovarian reserve [3].

In the setting of this increased use, the impact of the resulting supraphysiologic DHEA-S serum levels (following sulfation of DHEA) on sex steroid assays has not been fully evaluated. There have been reports that administration of DHEA augments progesterone production in women with low ovarian reserve undergoing ovarian tissue transplant [4, 5]. However, it is not clear if this is a physiologic phenomenon or if this could be due, at least in part, to assay artifact.

This distinction is of importance since late follicular rises in progesterone have been shown to adversely affect outcomes from fresh embryo transfers (ETs) [6, 7]. Many have advocated cryopreserving all embryos and avoiding fresh ET in the event that there is a premature rise in progesterone levels during an IVF cycle [8]. Given the data on premature progesterone levels, even modest alterations in this assay may impact clinical management and may be of importance when examining those on DHEA supplementation during an IVF cycle.

This study seeks to determine the relationship between DHEA supplementation and progesterone measurements to characterize the degree of interference with several commonly utilized commercial immunoassays.

Materials and methods

In order to determine DHEA-S interaction with commercial progesterone assays, the study was designed in two steps. First, DHEA-S control reagent standard with no progesterone present was run blinded on commercially available immunoassays for progesterone in order to determine if this would result in progesterone detection at varied levels. Second, serum pools from patients who were on DHEA and those who were not on DHEA supplementation were procured. DHEA-S was added to the pools from patients not on supplementation in order to determine if progesterone detection varied after the addition.

DHEA-S standard and progesterone assay

The first phase of the study focused on determining if the DHEA-S interacted with the progesterone assay from commonly utilized commercial vendors. In order to accomplish this, we utilized DHEA-S control reagents with no progesterone present measured in quadruplicate on three different immunoassay instruments: Advia Centaur (Siemens Healthcare Diagnostics, Inc., Tarrytown, NY), Immulite 2000 (Siemens Healthcare Diagnostics, Inc., Tarrytown, NY), and eCobas 411 (Roche Diagnostics USA, Indianapolis, IN). The three concentrations measured for DHEA-S were as follows: 0 μg/mL (control), 307 μg/mL (mid-range), and 722 μg/mL (high). This was analyzed for both DHEA-S and progesterone levels with just the DHEA-S standard present.

The mean values were compared using Pearson’s correlation.

DHEA-S standard in serum pools

The second phase of the study focused on the addition of physiologic ranges of DHEA-S to serum pools to mimic what is seen in patients undergoing IVF while on DHEA supplementation. Serum pools from 60 unique IVF patients’ serum were created. The serum was obtained with an IRB-approved protocol. There were six pooled serum samples: three pools from patients on DHEA supplementation and three pools from patients who were not on DHEA supplementation. In each of these two groups, the three pools were composed of patients whose serum fell into three different progesterone ranges: <1.5, 1.5–3.0, and >3 ng/mL.

Baseline DHEA-S and progesterone were measured in each of the six pooled serum samples. The mean level of DHEA-S in the mid-range progesterone pool from the patients who were on DHEA supplementation was used as the mid-point for addition of DHEA-S standard to the serum pools from patients without DHEA supplementation. Increasing amounts of DHEA-S were then added to the pooled serum from patients’ who were not on DHEA supplementation in the low progesterone group, and progesterone was measured in triplicate on three different immunoassay instruments: Advia Centaur (Siemens Healthcare Diagnostics, Inc., Tarrytown, NY), Immulite 2000 (Siemens Healthcare Diagnostics, Inc., Tarrytown, NY), and eCobas 411 (Roche Diagnostics USA, Indianapolis, IN).

The mean values were compared using Pearson’s correlation to determine if a relationship existed between DHEA-S levels and progesterone levels.

Results

DHEA-S standard and progesterone assay

When measuring the manufacturer’s DHEA-S controls, there was a linear increase in the progesterone detected. The values ranged from as low as 0.5 ng/mL in the blank control (no DHEA-S) to as high as 2.0 ng/mL in the high control (DHEA-S 722 μg/mL) in one assay system tested. This linear increase (p = 0.0002) in progesterone was seen despite the complete absence of progesterone in the sample being analyzed. These results are summarized in Fig. 1.

Fig. 1.

Measurement of the manufacturer’s DHEA-S controls showed a linear increase in the progesterone detected, ranging from 0.5 ng/mL without DHEA-S (CTRL) in the blank control (no DHEA-S) to as high as 2.0 ng/mL in the high control where DHEA-S was 722 μg/mL (High). This linear increase in progesterone was seen on all platforms despite the complete absence of progesterone in the sample being analyzed. Mean and SE bars shown

DHEA-S standard in serum pools

The results from the second phase are from the pooled serum samples from patients undergoing IVF. The six pooled serum results for DHEA-S and progesterone are shown in Table 1, clearly demonstrating that DHEA supplementation in fact results in increased serum DHEA-S levels. The DHEA-S levels for those not on DHEA supplementation are stable across the varied progesterone levels, while the DHEA-S levels for those on DHEA supplementation significantly increase across the progesterone pools. The DHEA-S level in the pooled serum from the mid-range progesterone group in patients on DHEA supplementation ranged from 283.5 to 392.0 μg/mL.

Table 1.

Pooled serum results from patients who are on oral DHEA supplementation and those that are not on oral DHEA supplementation with varied progesterone levels noted during their IVF cycle. Each pool consisted of 10 patients

		Low P4	Med P4	High P4
Oral DHEA supplementation	Progesterone assay (ng/mL)	0.88 (0.1),	1.41 (0.03),	3.18 (0.2),
		0.79–0.93	1.39–1.45	3.05–3.43
	DHEA-S assay (μg/mL)	283.5 (2.5),	322.8 (6.5),	392.0 (2.0),
		281.3–286.2	325.8–338.6	390.6–393.4
No oral DHEA supplementation	Progesterone assay (ng/mL)	0.55 (0.1),	1.53 (0.1),	3.55 (0.2),
		0.52–0.61	1.48–1.59	3.35–3.74
	DHEA-S assay (μg/mL)	142.0 (2.7),	131.2 (4.9),	143.8 (4.5),
		140.4–145.1	127.7–136.9	140.1–148.9

The samples were run in triplicate, and the mean, +/− SD, and ranges are shown

Then, the pooled serum sample from patients who were not taking DHEA supplementation had varied amounts of DHEA-S added to estimate a range which encompassed the variation of DHEA-S seen in those taking DHEA supplementation. No additional progesterone was added to the serum pools; the only difference in serum was the amount of DHEA-S. The DHEA-S levels ranged from 193.2 μg/mL in the baseline group to 317.4 μg/mL in group 1, 392.2 μg/mL in group 2, 487.5 μg/mL in group 3, and 504.9 μg/mL in group 4. The progesterone levels rose in all groups, summarized in Fig. 2. The Immulite 2000 (Siemens Healthcare Diagnostics, Inc., Tarrytown, NY) and eCobas 411 (Roche Diagnostics USA, Indianapolis, IN) assays had a linear relationship between DHEA-S added and the measured progesterone value (p < 0.05).

Fig. 2.

Varied progesterone assay results when DHEA-S was added to baseline serum pools. No progesterone was added in groups 1–4; the only variation is change in DHEA-S standard. DHEA-S levels varied from 193.2 μg/dL in the baseline group to 504.9 μg/dL in group 4. The progesterone assay showed increased progesterone in the serum sample despite no additional progesterone being present

Discussion

The data presented here show that DHEA-S can interfere with standard progesterone immunoassays used in clinical ART programs; in one assay system, this level reached as high as 2.0 ng/mL. These spurious progesterone assay elevations, in the complete absence of progesterone, may reach ranges shown to be detrimental in fresh IVF cycles, potentially altering clinical management. In the other assay systems, the level was below the 1.5 ng/mL cutoff often utilized, but this assay effect will be additive to the patients’ endogenous progesterone during an IVF cycle. This is an important finding as many programs implement premature progesterone rise as an indication for cryopreservation.

Of note, there have been other studies that have noted that serum progesterone levels during the follicular phase of IVF cycles were elevated in patients on DHEA supplementation [9]. In this case control study, the progesterone levels in the follicular phase of 15 patients was analyzed on stimulation day 5 and on the day of human chorionic gonadotropic administration. The serum progesterone levels were significantly elevated at both time points. In this instance, a physiologic basis for the findings was proposed. One explanation was that DHEA is one of the first compounds of the steroidogenic pathway and thus an increased substrate led to the increased progesterone identified [10, 11]. The other potential source cited was the increased number of follicles and an excess of granulosa cells which led to the increased progesterone levels [12]. In this circumstance, a progesterone assay effect was not considered.

The authors contacted the manufacturer of the assay standard. They were, at the time of this manuscripts creation, not aware of the interaction of DHEA-S standard with the progesterone assay. They noted that the interaction did not seem to be present with the physiologic levels of DHEA-S and that it was seen only at higher levels of standard and in those who were supplementing with DHEA during their IVF cycles. Although they could not suggest any immediate solution to the assay interaction, they suggested that, unless it was required for care during a treatment cycle that the causative interacting agent could be discontinued during the period of time during which the assay interaction may have an impact on clinical management.

The impact of this assay interaction is likely to be significant across clinics in the USA. According to data put out by the market research firm Research and Markets detailing the in vitro diagnostics forecast through 2020, the leading companies for testing are Roche, followed by Siemens and Abbott [13]. These companies control 80 % of the market share. The assay systems analyzed here were Roche and Siemens, the two largest distributors. There is not public information available from these companies regarding distribution of these assays across the USA; however, given their prevalence, it is likely to impact a number of fertility centers.

The data presented here suggest that an assay effect may be, at least in part, the reason for the observed progesterone rises noted in women who are taking DHEA supplementation. As many programs incorporate late follicular progesterone levels when making clinical decisions, such as timing of human chorionic gonadotropin administration and whether to cryopreserve all embryos, it is essential that each program identify the effect of DHEA supplementation on steroid assays. Further research is needed regarding the pharmacokinetics of DHEA supplementation to determine when it can be discontinued prior to an IVF stimulation so as not to interfere with assays or if the progesterone levels while on DHEA supplementation could be converted to reflect the true serum progesterone level. Alternatively, progesterone immunoassays may be developed that do not cross-react with the DHEA-S substrate.